English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Fairness is essential for machine learning systems deployed in high-stakeapplications. Among all fairness notions, individual fairness, deriving from aconsensus that `similar individuals should be treated similarly,' is a vitalnotion to describe fair treatment for individual cases. Previous studiestypically characterize individual fairness as a prediction-invariant problemwhen perturbing sensitive attributes on samples, and solve it byDistributionally Robust Optimization (DRO) paradigm. However, such adversarialperturbations along a direction covering sensitive information used in DRO donot consider the inherent feature correlations or innate data constraints,therefore could mislead the model to optimize at off-manifold and unrealisticsamples. In light of this drawback, in this paper, we propose to learn andgenerate antidote data that approximately follows the data distribution toremedy individual unfairness. These generated on-manifold antidote data can beused through a generic optimization procedure along with original trainingdata, resulting in a pure pre-processing approach to individual unfairness, orcan also fit well with the in-processing DRO paradigm. Through extensiveexperiments on multiple tabular datasets, we demonstrate our method resistsindividual unfairness at a minimal or zero cost to predictive utility comparedto baselines.

Learning Antidote Data to Individual Unfairness